CS-80 style VCO

[elektrouwe]

here is a schematic for a CS-80 style VCO. It is only a LTSPICE simulation until now. A guy asked for help on a german forum ( http://www.sequencer.de/synthesizer/viewtopic.php?f=13&t=78218 ) and I found it interesting to help him, because Yamaha used a closed loop VCO in their Chip. Although closed loop VCOs typically have better linearity than our "bread & butter discharge cap. with switch"-VCOs they are not popular in music electronics, because the rising ramp has a slope with fixed rise time and the amplitude is not constant over CV-range.
Rise time at low freq. does not play a role, but for frequencies in the kHz range , waveform changes towards an asymmetrical triangle. I guess that is the reason why Yamaha added a waveform shaper that simply masks the rising slope with a constant amplitude "punch" from the closed loop one-shot.
the short (30us ?) punch does not add audible harmonic content and, I guess again, is "grinded" away in the (bandwidth limited) signal path.
Please comment

---

CS-80_VCO.jpg
Description:	CS-80 style VCO schematics
Filesize:	219.33 KB
Viewed:	573 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

[steffensen]

I have nothing to add here, only that im glad someone finally picked this up. Will follow with interest!

[keyman2]

Looks deceptively simple...What would be required to make this a 'practical circuit' in terms of compensation for temperature or other variations which might alter tuning/performance?

[itist]

Hi,

if you want to listen to the VCO from elektrouwe i was the guy who asked for.
Here is a link to a short youtube Sound clip.

Here you can here it without Filter with my SPI Based 12Bit MIDI to Voltage converter.It sucks in the low frequecies but this is an issue of the converter. The VCO is stable.

http://youtu.be/k6iqJHWC2OY



http://youtu.be/N1N52uKgiHo

Its the CS Style VCO thru a 24 db Transistor ladder Filter.


Here you can here the VCO with a discrete DAC and its fine over the whole keyboard range.

http://youtu.be/uGv4gvDUYMU

I am searching for an VCO schematic collection espacially linear VCO

if someone has ideas or want me to do some measurements feel free..



Greetz


Oliver


http://www.youtube.com/user/DerITist
https://plus.google.com/+ItistDe
http://www.ititst.de
https://www.facebook.com/deriTist

[keyman2]

The ladder example sounds very saturated?

[itist]

[itist]

Hi,

i think one of Videos is not the CS Style VCO.

Today i have made some new measurements above the range of
125mV to 8 V in octave steps.

The voltage source is not my DAC to eliminate errors in the Voltage steps.

Voltage Frequency red Frequency yellow
0.125V 56.5Hz 56Hz
0.250V 112.5Hz 112Hz
0.5V 223.2Hz 223Hz
1V 439.9Hz 443Hz
2V 874.1Hz 874Hz
4V 1696Hz 1701Hz
8V 3226Hz 3231Hz

Measured with the Oscilloscop!!

all above 3v sounds detuned.

See Update!!



The circuit was powered wirh an standard linear Power supply no Lab supply.

Greetz

Oliver

---

125mV.jpg
Description:	125mV
Filesize:	86.38 KB
Viewed:	329 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

---

500mV.jpg
Description:	500mV
Filesize:	85.15 KB
Viewed:	325 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

---

1V.jpg
Description:	1V
Filesize:	84.79 KB
Viewed:	312 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

---

2V.jpg
Description:	2V
Filesize:	84.25 KB
Viewed:	320 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

---

4V.jpg
Description:	4V
Filesize:	87.96 KB
Viewed:	402 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

[itist]

After Uwe, the developer of the VCO, wrote me an e-mail where he asked me about the used OpAmp, i wrote him that i used an OP07 (bipolar), he told me to use jfet type. So I changed it to a LF356.

Here the results:

CV F
0.125V 49.8Hz
0.250V 99.01Hz
0.5V 197.4Hz
1V 393.8Hz
2V 784.9Hz
4V 1520Hz
8V 3140Hz

I think this is very linear.

I did a simple temperature check (hair dryer)

15C° 1000Hz
20C° 999Hz
30C° 993Hz
40C° 984Hz
46C° 977Hz

The circuit was build with standard metalfilm resitors with +- 50 ppm

Thank you Uwe.

[itist]

Hi,

Audio Demo CS Style VCO. Recorded direct to Soundcard no filters, no effects.



http://youtu.be/WoIxa03C_yE


The developer Uwe told me to check the used OpAmp. I used an bipolar instead of a Fet OpAmp
I changed the OpAmp from an OP07 to an LF 356 and the tracking Problems are gone...

Thanks Uwe.

Greetz
Oliver





http://itist.de/2013/10/cs-style-vco-core/

[Yipdeceiver]

I would love to make a +/- 12v version of this for my modular. What would I have to change to run this on +/-12v?

[elektrouwe]

[quote="Yipdeceiver"What would I have to change to run this on +/-12v?[/quote]
It will run @+-12V without problems, but a V/Hz slope change (mostly) from Iref = -V/R9 will happen.
R9 can be changed to a 12/15 lower value to get (almost) the same V/Hz slope as with +-15V.
It might be a good idea anyway to add a trimpot in series to R9 to compensate for other component tolerances.

[itist]

Hi,

i built two dual VCO boards of the CS style VCO created by Uwe.

http://itist.de/2014/06/dual-vco-board/

And a first Impressions Sound Demo
http://youtu.be/QQPX25NYFqs


Greetz Oliver

[elektrouwe]

[steffensen]

itist,
Would you care to share more info/details about the sync-isse you had and solved by using a buffer?
Never seen a solution like that before, only seen it solved by using precision voltage reference's and such. This makes me curious!

[itist]

In my first Impression Video I connected all outputs on the Diode D1 via a summing resistor Network
to an Opamp and then to the soundcard.

My assumption is that the outputs interfere with the trigger circuit (current return path??) and with the unity gain buffer this path is cut off.
Maybe it was only a kind of capacitive coupling. But with the buffer the sync issue has gone.

The circuit ist powered with a standard linear Voltage power supply but the control voltage is created with an LT1236.

For a linear VCO you need, especially in the lower key range, a precise Voltage. The lowest C is 125mV and the next C is 250mV
so each halftone is appr. 10 mV
I will do further testing but in my first tests the VCO seems to be linear over the Voltage Range from 125mV to 4mV.

You can follow the further development on my website:

www.itist.de

And please don't forget I am not an EE!
So my answers are maybe not correct!

Greetz
Oliver

[itist]

[steffensen]

itist,
Thanks for the info, always interesting to hear other peoples experiences!

Your thoughts makes sense, could very well be like you describe.

I have similar results with some other VCO's i have, but only been planning on solving it by precision voltage reference instead. Gonna give this a go too as there is no buffer implemented atm. (sorry to go off topic, getting carried away with interest)

Let me/us now how it goes!

[itist]

steffensen,

made a quick Video showing the effect.
http://youtu.be/zEfeB6HNb0w

Greetz

Oliver

[steffensen]

Excellent video man, thanks for doing it!
You got a nice workbench-area there for sure.
Hope i have some time on my hands to try this out soon, most curious.